[MINC-users] 4D files in register + Analyze obj map converter

[Soren Christensen]

Thanks Andrew.
I saw that manual entry but it makes more sense now!
I still can't figure out why register behaves the way it does, but I think
it is linked to that time dimension as the problem only occours on 4D files.
In any case it is purely purely a visual problem.

Cheers
Soren


On 11/15/06, Andrew Janke <a.janke at gmail.com> wrote:
>
> >  Is there a known issue when displaying 4D files in register?  (I
> realize
> > that just the first frame is visualized)
> > I am trying to make some 4D DTI datasets in MINC format "nifti friendly"
> by
> > reshaping to get positive spatial orientation of data as descibed in
> >
> http://brainbody.nottingham.ac.uk/users/wiki/pmwiki.php?n=MincUsersGroup.FAQDataFormats
> > .
> > When I look at the original minc and the reshaped minc in register,
> using
> > the crosshair in sync mode shows that the images are spatially
> identical.
> > However, the composite image does not show overlay of the two
> images.  Is
> > there an explanation for this?
>
> There is no issue that I am aware of.
>
> > Also, what is the explanation for the dimsizes of -1 in order to achive
> > flipping of the data (below)?
> >
> > mincreshape +direction -dimsize xspace=-1 -dimsize yspace=-1 -dimsize
> > zspace=-1 -dimorder zspace,yspace,xspace infile.mnc outfile.mnc
>
> You will find this hidden functionality deep within the mincreshape
> manual page. Essentially, what this does is allow for all dimensions
> within the image to be flipped to a positive step. Irrespective as to
> whether they are image or slice dimensions. The default is to only
> flip image dimensions, that is the dimension that is the slowest
> varying within your file.
>
> > If anyone should ever need (and it is not already available?) it I have
> a
> > little perl script and a binary/c++ source to read Analyze object maps.
> The
> > binary binarizes the Analyze objmap (run length encoded) and this can
> then
> > be  converted to minc.
> > A perl script converts the binarized objmap to MINC given a minc
> template
> > file. There are options for -flip xyz and dimorder as the Analyze objmap
> is
> > has a very modest header and spatial information depends on the file on
> > which it was drawn.
>
> By all means, I am always more than interested in new converters to
> add to the MINC Toolkit. In this case, it will probably be best if you
> e-mail it to myself directly.
>
> Thanks
>
>
> --
> Andrew Janke   (a.janke at gmail.com || http://a.janke.googlepages.com/)
> Canada->Montreal                                   Cell: +1 (514) 924 2012
> _______________________________________________
> MINC-users at bic.mni.mcgill.ca
> http://www.bic.mni.mcgill.ca/mailman/listinfo/minc-users
>